LWN headline: Blame Fedora = High Praise

[Stephen John Smoogen]

On 5/1/07, Matt Domsch <matt at domsch.com> wrote:
> Like it or not, Operating Systems and software run on hardware.
>

I am agreeing with Matt on this.

First doing this breaks the "Work with Upstream" on this.
Second it doesnt seem to follow the "Work as a subteam first, see how
it works, then integrate with mainline."

I would say that the first step would be for Alexandre Oliva and
similar minded people to create a SIG and then build a working
kernel+tools around Fedora 7 final to get a working design done. When
you have working code and can show how people will work with it, then
you can work on getting integrated or know that it isnt going to work
at all within Fedora due to non-technical problems

Second when you have a working model, you can get mindshare to help
work changing upstream that it shouldnt have any firmware in the
kernel. Unless you can really get upstream to buy into this.. then it
will not really fly with maintenance of the Fedora.

Third you are going to need to get some sort of system builder aligned
with your goal.. as getting a system for people that is completely
hackable is not easy or cheap at this time. When you can get a
marketable reason that this is better than doing it otherwise.. you
can change minds and buying decisions.

All of these will take time and effort... you can get 1 done in a
year, 2 and 3 will take 4-5 years of sustained effort and learning
from failures.


-- 
Stephen J Smoogen. -- CSIRT/Linux System Administrator
How far that little candle throws his beams! So shines a good deed
in a naughty world. = Shakespeare. "The Merchant of Venice"